Limits...
Training-induced compensation versus magnification of individual differences in memory performance.

Lövdén M, Brehmer Y, Li SC, Lindenberger U - Front Hum Neurosci (2012)

Bottom Line: Initial mnemonic instructions reduced between-person differences in memory performance, whereas further practice after instruction magnified between-person differences.We conclude that strategy instruction compensates for inefficient processing among the initially less able.In contrast, continued practice magnifies ability-based between-person differences by uncovering individual differences in memory plasticity.

View Article: PubMed Central - PubMed

Affiliation: Center for Lifespan Psychology, Max Planck Institute for Human Development Berlin, Germany.

ABSTRACT
Do individuals with higher levels of task-relevant cognitive resources gain more from training, or do they gain less? For episodic memory, empirical evidence is mixed. Here, we revisit this issue by applying structural equation models for capturing individual differences in change to data from 108 participants aged 9-12, 20-25, and 65-78 years. Participants learned and practiced an imagery-based mnemonic to encode and retrieve words by location cues. Initial mnemonic instructions reduced between-person differences in memory performance, whereas further practice after instruction magnified between-person differences. We conclude that strategy instruction compensates for inefficient processing among the initially less able. In contrast, continued practice magnifies ability-based between-person differences by uncovering individual differences in memory plasticity.

No MeSH data available.


Related in: MedlinePlus

Graphical representation of the confirmatory factors model (A) and the latent differences score model (B) used to estimate gains from mnemonic instruction (baseline plasticity). Observed variables are represented by squares, latent variables by circles, regression weights by one-headed arrows, and variances and covariances by two-headed arrows. The triangle indicates means. Unlabeled parameters are fixed to 1. BP, baseline performance; POST, post-instruction performance; l, list.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3351801&req=5

Figure 1: Graphical representation of the confirmatory factors model (A) and the latent differences score model (B) used to estimate gains from mnemonic instruction (baseline plasticity). Observed variables are represented by squares, latent variables by circles, regression weights by one-headed arrows, and variances and covariances by two-headed arrows. The triangle indicates means. Unlabeled parameters are fixed to 1. BP, baseline performance; POST, post-instruction performance; l, list.

Mentions: To analyze instruction gains (i.e., the difference between baseline and post-instruction assessments; see Table 2) we fitted a confirmatory two-factor model to the data from the baseline assessment and the post-instruction assessment (see Figure 1A). That is, we assumed a latent unobserved variable representing an individual's latent error-free baseline performance score (BP) before introduction to the mnemonic technique and a latent variable representing an individual's score after instruction (Post). The latent BP score is defined as a unit-weighted factor of two observed variables [list 3 (l3) and list 4 (l4)], representing performance on the first and second lists using landmark cues in the baseline assessment (the first two lists had numbers as cues), respectively. The latent post-score is defined as a unit-weighted factor of two other observed variables (l7 and l8), representing performance on the third and fourth lists presented to participants in the post-instruction assessment. The reason for including only two lists from the post-instruction assessment was to match the list-order of the lists tapping baseline performance. We simultaneously and freely estimate the error variances (σ2e3, σ2e4, σ2e7, and σ2e8), the autocovariances between the errors (ρe3,e7 and ρe4, e8), and the mean difference between the lists used as indicators of baseline and post-instruction performance (μlistdiff). Of particular interest, we simultaneously estimate the mean of baseline performance (μBP), interindividual differences in baseline performance (σBP), the mean of the latent post-instruction performance (μpost), interindividual differences in post-instruction performance (σpost), and the correlation between baseline performance and post-instruction performance (ρBP, post). We also included the cognitive composites of perceptual speed, episodic memory, reasoning, and verbal knowledge as observed variables, and allowed these to freely covary among themselves and with latent baseline performance and post-instruction performance (not shown in Figure 1A)1. In order to compare the estimates across age groups, we estimated this model as a multigroup model (children, younger adults, and older adults). In the starting model, no across-group constraints were applied. With this model, we can inspect the standard deviations of the latent factors, baseline performance, and post-instruction performance, and test for the effects of training on between-person differences expected from the compensation and magnification views.


Training-induced compensation versus magnification of individual differences in memory performance.

Lövdén M, Brehmer Y, Li SC, Lindenberger U - Front Hum Neurosci (2012)

Graphical representation of the confirmatory factors model (A) and the latent differences score model (B) used to estimate gains from mnemonic instruction (baseline plasticity). Observed variables are represented by squares, latent variables by circles, regression weights by one-headed arrows, and variances and covariances by two-headed arrows. The triangle indicates means. Unlabeled parameters are fixed to 1. BP, baseline performance; POST, post-instruction performance; l, list.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3351801&req=5

Figure 1: Graphical representation of the confirmatory factors model (A) and the latent differences score model (B) used to estimate gains from mnemonic instruction (baseline plasticity). Observed variables are represented by squares, latent variables by circles, regression weights by one-headed arrows, and variances and covariances by two-headed arrows. The triangle indicates means. Unlabeled parameters are fixed to 1. BP, baseline performance; POST, post-instruction performance; l, list.
Mentions: To analyze instruction gains (i.e., the difference between baseline and post-instruction assessments; see Table 2) we fitted a confirmatory two-factor model to the data from the baseline assessment and the post-instruction assessment (see Figure 1A). That is, we assumed a latent unobserved variable representing an individual's latent error-free baseline performance score (BP) before introduction to the mnemonic technique and a latent variable representing an individual's score after instruction (Post). The latent BP score is defined as a unit-weighted factor of two observed variables [list 3 (l3) and list 4 (l4)], representing performance on the first and second lists using landmark cues in the baseline assessment (the first two lists had numbers as cues), respectively. The latent post-score is defined as a unit-weighted factor of two other observed variables (l7 and l8), representing performance on the third and fourth lists presented to participants in the post-instruction assessment. The reason for including only two lists from the post-instruction assessment was to match the list-order of the lists tapping baseline performance. We simultaneously and freely estimate the error variances (σ2e3, σ2e4, σ2e7, and σ2e8), the autocovariances between the errors (ρe3,e7 and ρe4, e8), and the mean difference between the lists used as indicators of baseline and post-instruction performance (μlistdiff). Of particular interest, we simultaneously estimate the mean of baseline performance (μBP), interindividual differences in baseline performance (σBP), the mean of the latent post-instruction performance (μpost), interindividual differences in post-instruction performance (σpost), and the correlation between baseline performance and post-instruction performance (ρBP, post). We also included the cognitive composites of perceptual speed, episodic memory, reasoning, and verbal knowledge as observed variables, and allowed these to freely covary among themselves and with latent baseline performance and post-instruction performance (not shown in Figure 1A)1. In order to compare the estimates across age groups, we estimated this model as a multigroup model (children, younger adults, and older adults). In the starting model, no across-group constraints were applied. With this model, we can inspect the standard deviations of the latent factors, baseline performance, and post-instruction performance, and test for the effects of training on between-person differences expected from the compensation and magnification views.

Bottom Line: Initial mnemonic instructions reduced between-person differences in memory performance, whereas further practice after instruction magnified between-person differences.We conclude that strategy instruction compensates for inefficient processing among the initially less able.In contrast, continued practice magnifies ability-based between-person differences by uncovering individual differences in memory plasticity.

View Article: PubMed Central - PubMed

Affiliation: Center for Lifespan Psychology, Max Planck Institute for Human Development Berlin, Germany.

ABSTRACT
Do individuals with higher levels of task-relevant cognitive resources gain more from training, or do they gain less? For episodic memory, empirical evidence is mixed. Here, we revisit this issue by applying structural equation models for capturing individual differences in change to data from 108 participants aged 9-12, 20-25, and 65-78 years. Participants learned and practiced an imagery-based mnemonic to encode and retrieve words by location cues. Initial mnemonic instructions reduced between-person differences in memory performance, whereas further practice after instruction magnified between-person differences. We conclude that strategy instruction compensates for inefficient processing among the initially less able. In contrast, continued practice magnifies ability-based between-person differences by uncovering individual differences in memory plasticity.

No MeSH data available.


Related in: MedlinePlus